h tt p : / / w w w . b j m i c r o b i o l . c o m . b r /
Biotechnology
and
Industrial
Microbiology
Selection
of
starter
cultures
for
the
production
of
sour
cassava
starch
in
a
pilot-scale
fermentation
process
Fernanda
Corrêa
Leal
Penido
a,∗,
Fernanda
Barbosa
Piló
b,
Sávio
Henrique
de
Cicco
Sandes
c,
Álvaro
Cantini
Nunes
c,
Gecernir
Colen
a,
Evelyn
de
Souza
Oliveira
a,
Carlos
Augusto
Rosa
b,
Inayara
Cristina
Alves
Lacerda
aaUniversidadeFederaldeMinasGerais,FaculdadedeFarmácia,DepartamentodeAlimentos,BeloHorizonte,MG,Brazil
bUniversidadeFederaldeMinasGerais,InstitutodeCiênciasBiológicas,DepartamentodeMicrobiologia,BeloHorizonte,MG,Brazil cUniversidadeFederaldeMinasGerais,InstitutodeCiênciasBiológicas,DepartamentodeBiologiaGeral,BeloHorizonte,MG,Brazil
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:Received18May2017 Accepted5February2018 Availableonline28February2018 AssociateEditor:SolangeI. Mussatto
Keywords:
Lacticacidbacteria Yeasts
Startercultures Fermentation Bakeryproducts
a
b
s
t
r
a
c
t
Sourcassavastarch(Polvilhoazedo)isobtainedfromaspontaneousfermentationconducted bymicroorganismsfromrawmaterialsandfermentationtanks.Thisproductis tradition-allyusedinthebakingindustryforthemanufactureofbiscuitsandBraziliancheesebreads. However,theendoffermentationisevaluatedempirically,andtheprocessoccurswithout standardization,whichresultsinproductsofinconsistentquality.Predominantmicrobiota fromacassavaflourmanufacturerwasisolatedinordertoselectstarterculturesforthe productionofsourcassavastarchinapilot-scalefermentationprocess.Lacticacid bacte-riaand yeasts wereisolated,enumerated andgroupedbyRestrictionFragment Length Polymorphism,andPCRfingerprinting,respectively.Oneisolateofeachmolecularprofile wasidentifiedbysequencingoftherRNAgene.LABwereprevalentthroughouttheentire process.Lactobacillusbrevis(21.5%),whichproducedthehighestvaluesofacidity,and Lac-tobacillusplantarum(13.9%)wereamongthemostfrequentspecies.Pichiascutulata(52.2%) wastheprevalentyeastandshowedamylolyticactivity.Theaforementionedspecieswere testedassingleandmixedstarterculturesinapilot-scalefermentationprocessfor28days.
L.plantarumexhibitedbetterperformanceasastarterculture,whichsuggestsitspotential fortheproductionofsourcassavastarch.
©2018SociedadeBrasileiradeMicrobiologia.PublishedbyElsevierEditoraLtda.Thisis anopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/
licenses/by-nc-nd/4.0/).
∗ Correspondingauthor.
E-mail:fclpenido@gmail.com(F.C.Penido).
https://doi.org/10.1016/j.bjm.2018.02.001
1517-8382/©2018SociedadeBrasileiradeMicrobiologia.PublishedbyElsevierEditoraLtda.ThisisanopenaccessarticleundertheCC BY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Introduction
Becausecassava(ManihotesculentaCRANTZ)hasahighstarch content (approximately 80%), it is an important source of carbohydratesthatcanbesoldfreshorprocessedintoa vari-ety ofvalue-added products; cassava ischaracterized as a multipurposecrop.1In2014,theestimatedBraziliancassava
productionwas23milliontons.2Althoughwidelyconsumed,
cassavahaslimitationsduetoitsperishability,toxicityand lowproteincontent.3Therootofcassavacontainscyanogenic
glycosidesthatactasdefencesubstancesthroughtherelease ofhydrogencyanide,whichisresponsibleforitstoxicity.The traditional fermentations of cassava are quite suitable for thedetoxification,preservationanddevelopmentofproducts withdesirableviscoelastictexture.4Lacticfermentationnot
onlyextendstheshelflifeofthisrootbutalsodecreasesits toxicity.5
Naturalfermentationofwetstarchextractedfromcassava rootisatraditionaltechnologywidelyusedinLatin Amer-ica.Sweet cassava starchand sour cassavastarch undergo thesameprocessofstarchextraction,butdifferin fermen-tationtime(fromtwotosevendaysandfrom20to70days, respectively) and therefore have different levels of acidity, maximum of1% ofacidity and 5% ofacidity, respectively. Sourcassavastarchstandsoutfromsweetcassavastarchand otherflourinthepreparationofbakeryproductsforitsunique expansioncapacity,withouttheadditionofbakingsodaand in the absence of gluten. Sour cassava starch is a typical Brazilianfoodgenerallyproducedbysmallandmedium-sized ruralindustries.6Assourcassavastarchisessentially
hand-crafted,eventhoughitisproducedinmoderncassavaflour manufacturers, it still has heterogeneous physicochemical andsensoryquality.Thesuccessionofmicroorganismsfrom rawmaterialsandfermentationtanksoccursnaturally dur-ingcassavafermentationandresultsinamicrobiotawitha prevalenceoflacticacidbacteria(LAB),suchasLactobacillus plantarum,Lactobacillusfermentum,Lactobacillusbrevisand Leu-conostocmesenteroides.7–10 Inthe productionofsourcassava
starchinBrazil,Lactobacillusoccursinassociationwithyeasts, suchasGalactomycesgeothricumandIssatchenkia(nowPichia)
sp.6,11
Thewiderange andcomplexity ofcassavaspontaneous fermentationmicrobiotaarethemainfactorsresponsiblefor thelackofhomogeneityandlowproductquality.Theuseof selectedstrainsisanimportantalternativebecauseitprovides lessvariationinthecontentofchemicalcompounds,shorter fermentationtime,higheryieldandsensorialquality.12,13The
LAB,whichareusedasnaturalorselectedstarterculturesin fermentedfoods,areabletoacidifyandenhancetheflavor. Furthermore, the LAB can protect food from the develop-ment of pathogens due to the formation of antimicrobial compounds.14,15
AlthoughsourcassavastarchiswidelyconsumedinLatin Americaandgained prominenceinthe preparationof bak-eryproductswithalowlevelortheabsenceofgluten,there are fewstudieson itsfermentationprocess.Therefore, the studyofstarterculturescontributessignificantlytothe under-standingandoptimizationofsourcassavastarchproduction. Thepresentworkaimedtoselectthestartercultureswiththe
appropriatecharacteristicsfortheproductionofsourcassava starchinapilot-scalefermentationprocess.
Materials
and
methods
Collectionofsamplesfromcassavaflourmanufacturer
Atotalof16samplesof100geachwerecollectedfroma cas-savaflourmanufacturerinthemunicipalityofFormiga,Minas Gerais(MG)state,Brazil,ondays0,5,12,19,26,33and40of a56-dayspontaneousfermentationfromafermentationtank withusablecapacityof16,000L.Thesesampleswere trans-portedtothelaboratoriesofFoodMicrobiologyandIndustrial Microbiology and Biocatalysis(Faculdadede Farmácia, Uni-versidade Federal de Minas Gerais, MG,Brazil) on ice and processedwithin24h.Theprocessingconsistedofweighing twenty-fivegramsofeachsampleinsterileflasks,dilutedin 225mLof1gL−1peptonewater,andpreparingserialdecimal dilutions.16
Identificationoflacticacidbacteriafoundonsamplesfrom cassavaflourmanufacturer
AppropriatedecimaldilutionswerespreadondeMan,Rogosa andSharpe(MRS;Acumedia,Lansing,MI,USA)agar contain-ing0.1gL−1cycloheximide16andincubatedinanaerobicjars
of 2.5L (Permution, Curitiba, Brazil)at 37◦C for 48h. After growth,onecolonyofeachmorphotypewascountedand puri-fiedforlateridentification.EachisolatewasGramstainedand thensubjectedtothecatalasetest.17
DNA was extracted with an adaptation of the method described by Hoffman and Winston.18 The colonies
previ-ously grown on MRS agar were resuspended in 100L of Tris–EDTA (TE).Then, 100Lofphenol–chloroform–isoamyl alcohol(25:24:1) and0.3gofglassbeadswere addedtothe suspension. Tubes containing this mixture were homoge-nized by a vortex shaker (QL-901, Biomixer, Santa Clara, CA, USA) for three to4min and centrifuged at 13,000rpm for 5min(Eppendorf, Hamburg, Hamburg, Germany).After that,thesupernatantwastransferredtoanothertube.Then, a volume of 960mLL−1 ethanol corresponding to the vol-ume ofthe supernatant recovered was added to the tube. Thetubes were homogenizedbyinversionand centrifuged at 13,000rpm for 2min. The liquid phase was discarded, the tubes were driedovernight,and the DNA resuspended in 50LofTE. TheDNA concentration wasdetermined by NanoDropND1000Spectrophotometer(NanoDropProducts, Wilmington, DE,USA). TheDNA oflactic acid bacteriawas subjectedtoPCRamplificationofthe16SrRNAgeneusingthe primers27F(5-AGAGTTTGATCCTGGCTCAG-3)and1492R(5 -GGTTACCTTGTTACGACTT-3).19AllLABisolatesweregrouped
byRestrictionFragmentLengthPolymorphism(RFLP)by diges-tionwithrestrictionenzymesMspI,HinfIandHaeIII(Promega Corporation, Madison, WI, USA) according to the modified methodology of Brightwell et al.20 For the digestion
reac-tion,2Lof10×buffer,2Lofbovineserumalbumin(BSA) onlyfortheMspIenzyme,1Lofenzyme,DNA≤1500ng/L and water q.s.p. 20L. Thetubes were incubated at 37◦C for 3h. Therestriction fragments obtainedwere separated
by2%agarosegelelectrophoresis(Pronadisa,Spain)in0.5% TBE buffer at 100V. Thegels were stained withGelRedTM solution(Biotium,USA)andvisualizedunderultravioletlight (UV)byanimagecapturesystem(VilberLourmat,France).An isolateofeachdifferentmolecularRFLPprofilewasselected andsubjectedto16SrRNAsequenceanalysis.19Sampleswere
sequencedbycapillaryelectrophoresisinABI3130equipment usingPOP7polymerandBigDyev3.1(MyleusBiotechnology, BeloHorizonte,MG,Brazil).Thesequenceswereassembled, editedandalignedwiththeprogramMEGA6.21Thesequences
obtainedwerecomparedwiththoseincludedintheGenBank databaseusingtheBasicLocalAlignmentSearchTool(BLAST
athttp://www.ncbi.nlm.nih.gov).
DifferentiationofspeciesoftheL.plantarumgroup
Alltheisolatesthatwerepresumablyidentifiedby16SrRNA sequenceanalysisasbelongingtotheL.plantarumgroupwere subjected to Multiplex PCR Assay with recA Gene-Derived Primers. This analysis allows the separation of the three closelyrelatedspeciesfromtheL.plantarumgroupby com-parisonofthesizeoftheiramplicons:318bpforL.plantarum,
218bpforLactobacilluspentosus and 107bpforLactobacillus paraplantarum.22 L. plantarum ATCC1 4917, L. paraplantarum
DSM10667andL.pentosusATCC8041wereusedascontrols.
Identificationofyeastsfoundonsamplesfromcassava flourmanufacturer
Portionsofappropriatedecimaldilutionswere spread onto yeastextract–maltextract(YM;Acumedia,Lansing,MI,USA) agarcontaining0.2gL−1chloramphenicolandincubatedfor 48hat25◦Cunderaerobicconditions.6 Onecolonyofeach
differentmorphotypewascountedandpurifiedforlater iden-tification.
ThetotalDNA ofeach yeast isolatewas extracted with an adaptation of the method described by da Silva-Filho et al.23 The colonies previously grown on YM agar were resuspendedin100Loflysisbuffer(Tris–HCl– trishydrox-ymethylaminomethane–0.05M,0.05MEDTA,0.1MNaCland 10gL−1 SDS –sodium dodecyl sulfate). Thetubes contain-ingthesuspensionwereincubatedinawaterbathat65◦C for35min.Then,100Lofphenol–chloroform–isoamyl alco-hol(25:24:1)wereaddedtothesuspension.Tubescontaining thismixturewerehomogenizedbyavortexshaker(QL-901, Biomixer,SantaClara,CA,USA)for3minandcentrifugedfor 15min(Eppendorf,Hamburg,Hamburg,Germany).Afterthat, thesupernatant wastransferredto anothertube, towhich wasadded100Lofcold700mLL−1ethanol.Thetubeswere homogenizedbyinversionandcentrifugedat13,000rpmfor 3min(Eppendorf,Hamburg,Hamburg,Germany).Theliquid phasewasdiscardedandthetubesweredriedovernight.The DNAwasresuspendedin100LofTE,andstoredinafreezerat −20◦C.TheDNAconcentrationwasdeterminedbyNanoDrop
ND1000Spectrophotometer(NanoDrop Products, Wilming-ton,DE,USA).
Allyeastisolatesweregroupedbytheirmolecularprofiles usingthePCRfingerprintingtechniquewiththemicrosatellite primer (GTG)5 (5-GTGGTGGTGGTGGTG-3).24 The PCR
mix-ture contained2.5Lof 10× buffer, 1Lof dNTPs0.1mM,
1.5LofMgCl2,2Loftheprimer(GTG)510mol−1
(Invitro-gen,Carlsbad,CA, USA),0.2LofTaqDNA polymerase1U and 1LofDNAina totalvolume of25L.ThePCR reac-tionwasperformedontheMastercycler®Proandshowedthe followingconditions:initialdenaturationat94◦C for2min, 40cyclesofdenaturationat95◦Cfor45s,annealingat50◦C for1minandextensionat72◦Cfor1min,followedbyfinal extensionat72◦Cfor6min.Onerepresentativeofeach dif-ferentmolecularprofilewasidentifiedbysequencingofthe D1/D2 domains of the large subunit of rRNA gene using theprimersNL1(5-GCATATCAATAAGCGGAGGAAAAG-3)and NL4(5-GGTCCGTGTTTCAAGACGG-3)accordingtoLachance etal.25ThePCRmixturecontained5Lof10×buffer,1Lof
dNTPs0.05mM,3LofMgCl21.5mM,1LoftheprimerNL1
10mol−1(Invitrogen,Carlsbad,CA,USA),1Loftheprimer
NL410mol−1(Invitrogen,Carlsbad,CA,USA),0.2LofTaq DNApolymerase1Uand1LofDNAinatotalvolumeof50L. Thereaction was performedon theMastercycler® Pro and showedthefollowingconditions:initialdenaturationat95◦C for2min,35cyclesofdenaturationat95◦Cfor15s,annealing at54◦Cfor25s,andextensionat72◦Cfor20s,followedbyfinal extensionat72◦Cfor10min.Thesamplesweresequenced bycapillaryelectrophoresisinABI3130equipmentusingPOP7 polymerandBigDyev3.1.Sequenceanalyseswereperformed asdescribedabove.
Evaluationofsamplesfromcassavaflourmanufacturer
Total titratable acidity (TTA) and pH of samples were determined.26Thefinalproductfromthefermentationtank
was submitted to microbiological analyses to search for
Bacillus cereus,thermotolerantcoliforms andSalmonellaspp. accordingtoBrazilianlegislation.27
Screeningforselectionofstrainstobetestedassingleor mixedstarterculture
All thestrainsofLABspeciesisolatedwere testedin tripli-cateforstarchdegradationonplateswithMRSagarcontaining 20gL−1 solublestarch28 andallthe strainsofyeastspecies
isolated,onplateswithYMagarcontaining20gL−1 soluble starch.29 After growth, the revelation was performed with
iodinesolutioninordertoviewstarchhydrolysishalos.Total acid productionwas evaluated in100mLofbroth contain-ing20gL−1(non-fermented)cassavastarch,10gL−1 glucose and5gL−1ofbeefextractwith24and48h.Aliquotsof10mL ofbrothwereusedforTTAmeasurement.26Thestrainsthat
showed starchdegradation halosandhighervaluesofTTA weresubsequentlytestedasstartercultures.
Useofstarterculturesinapilot-scalefermentationprocess
Single culturesofselected LAB strains and mixedcultures oftheseinassociationwiththe yeastwere inoculatedinto 100mLofculturemediumcontaining20gL−1(non-fermented) cassavastarch,10gL−1 glucoseand5gL−1 beefextractand thenincubatedatroomtemperaturefor24–48h.Theentire fermentedbrothwasusedtoinoculate500mLofthesame mediumfor24–48h.Theresultingbrothwas usedto inoc-ulate 5Lof a mediumcontaining 100gL−1 cassava starch,
for 28 days. The 5-L fermentations were inoculated with 8log10CFU/mLofLABand/or6log10CFU/mLofyeasts.Assays
wereperformedwithoutaeration.
Sampleswereweekly collectedintriplicatefor determi-nation of pH and TTA26 as well as for verification of the
viabilityofcultures.Themonitoringofstartercultureswas donebyisolationandpurificationofthreecoloniesperweek ofeachLABoryeastthatresembledmorphologicallytothe starterculturesselectedandperformedwithmolecular meth-odsdescribedabove,andconductedbycomparisonofprofiles obtainedwithprofilesalready known.Afterthe endof fer-mentation,thematerialwasdriedatroomtemperaturefor 5days,packedandstoredunderrefrigerationuntil comple-tionofmeasurementsofpHandTTA26;expansioncapacity30;
andmicrobiologicalanalysis,asdescribedabove.The expan-sioncapacityassaywascarriedoutaccordingtotheprocedure proposed byTropicalRoots Center.Fortymilliliters of boil-ingdistilledwaterwereaddedto50gofcassavastarch.Each masswasmodeledformakingroundcookiesofapproximately 10geach.Thediametersofthecookiesweremeasuredwith auniversalpachymeter(Series530,Mitutoyo)beforeandafter takingthemtotheovenforaperiodof20minat220◦C.The expansioncapacitywas calculatedbythe ratio ofaverages ofthe finaldiameter(afterbaking)and theinitialdiameter (beforebaking).
Dataanalysis
The data obtained were subjected to analysis of variance (ANOVA) and significant differences between means were determinedbyTukey’stestwitha0.05significancelevel.
Results
Identificationofmicroorganismsfromcassavaflour manufacturer
LAB were present in all samples during a 56-day cassava fermentation with higher counts (5.8–7.9log10CFU/g) than
those obtained for yeasts (1.7–7.8log10CFU/g), which were
found from day 0 to 26. The 79 isolates of LAB identi-fiedwereGram-positive, catalasenegativeandmostly rods (83%).ThebacterialisolateswereefficientlygroupedbyRFLP
(Figs.1and 2)intwelvedifferent species.Sevenspeciesof
Lactobacilluswereisolated:L.brevis (n=17,21.5%),L. fermen-tum (n=12, 15.2%), L. plantarum (n=11, 13.9%), Lactobacillus casei/Lactobacillusparacasei(n=7,8.9%),Lactobacillusharbinensis
(n=5, 6.3%), Lactobacillus parabuchneri (n=4, 5.0%) and Lac-tobacillus ghanensis (n=1, 1.3%). Enterococcus faecium (n=13, 16.5%)wasalsoisolatedand,inminorproportions,Weisella cibaria(n=3,3.8%),Lactococcusgarvieae(n=3,3.8%),Lactococcus lactissubsp.lactis(n=2,2.5%)andL.mesenteroides(n=1,1.3%). ThemostfrequentisolateswereL.brevis,isolatedfromsixout ofsevensamplecollections(5.8–7.5log10CFU/g),andL.
plan-tarum,isolatedinfivesamplecollections(5.4–6.4log10CFU/g).
Alloftheelevenisolatesthatwerepresumablyidentified by16SrRNAsequenceanalysisasbelongingtotheL.plantarum
groupwereposteriorlyconfirmedasspeciesofL.plantarumby MultiplexPCRAssayusingrecAGene-DerivedPrimers.
The23yeastisolateswereefficientlygroupedbyanalysisof theirmolecularprofilesobtainedbyPCRfingerprinting(Fig.3). ThemostcommonlyisolatedyeastspecieswerePichia scutu-lata(n=12,52.2%)andKazachstaniaexigua(n=4,17.4%),both foundfromday0to12(>4log10CFU/g).Candidahumilis(n=3,
13.0%),Geotrichumfragrans(n=3,13.0%)andCandidaethanolica
(n=1,4.4%)werealsoisolated.
Evaluationofsamplesfromcassavaflourmanufacturer
The pH values decreased and there was a consequent increase in TTA (Table 1). Bacillus spp. were found in the finalproductsample,withcountsof9.33×102CFU/g,which
is a valuelower than the tolerancepermitted byBrazilian law(3.00×103CFU/g).31Neitherthermotolerantcoliformsnor
Salmonellaspp.weredetected.
Screeningforselectionofstrainstobetestedassingleor mixedstarterculture
Noneofthe82LABstrainsisolatedexhibitedamylolytic activ-ity.However,outofthe33yeaststrainsisolatedalltheones belongingtothespeciesP.scutulata,C.humilisandC.ethanolica
showedpositiveresults(datanotshown)forstarch degrada-tion (indicatedbyaclearzone ofinhibition>1mmaround colonies).TTAvaluesevaluatedforthepredominant micro-biotaweregreaterat48hthanat24hoffermentationformost oftheLAB.ThebacteriawithhigheracidityvalueswereL. bre-visandL.garvieae.Tables2and3showtheresultsofTTAfor thebeststrainisolatedofeachspecies.
OneisolateofL.brevisandoneisolateofL.plantarumwere selectedtobetestedasthestarterculturesbecausetheywere amongthepredominantLABthathadhighervaluesofacidity. Theisolateofyeastchosentobetestedasstarterculturewas onestrainofP.scutulata,thepredominantyeastspecies,which exhibitedamylolyticactivity.Moreover,P.scutulataferments glucose,whichisimportantfortheproductionofsecondary metabolitessuchasvolatilecompoundsresponsibleforthe characteristicaromaofsour cassavastarch.In the present study,wetestedfourstartercultures:L.brevis;L.plantarum; L.brevisinamixedculturewithP.scutulata;andL.plantarum
inamixedculturewithP.scutulata.
Useofstarterculturesforsourcassavastarchprocessing
MonitoringofTTAandpHduringthepilot-scalefermentation processrevealedthatvaluesobtainedfordifferentstarter cul-turesdidnotdiffer(Fig.4).TTArangedfrom0.14%to0.71%. The pHrangedfrom 5.69to 3.29.ANOVA showed that the extentoffermentationsignificantlychangedtheTTAandpH ofcassavastarch.
The starter cultures used were able to prevail during cassava pilot-scale fermentation; therefore, all the isolates selectedforidentificationdidcorrespondtothestarter cul-turesinoculated.TheLABwasobservedtoremainin fermen-tationduring28days.ItwasobservedthatP.scutulatashowed amaximumcount,approximately7log10CFU/g,within7days
offermentation,andthegrowthdecreasedreachingcounts between3and4log10CFU/gwith14days.Thenumberofviable
10000 bp
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
1000 bp 500 bp 250 bp
Fig.1–Digestionprofileoflacticacidbacteriaisolatedfromcassavafermentationinacassavaflourmanufacturer.Lanes:1: 1kbDNAladder;2–4:Lactobacillusbrevisprofile(2-MspI,3-HaeIII,4-HinfI);5–7:Lactobacillusfermentumprofile(5-MspI, 6-HaeIII,7-HinfI);8–10:Lactobacillusplantarumprofile(8-MspI,9-HaeIII,10-HinfI);11–13:Lactobacilluscasei/L.paracaseiprofile (11-MspI,12-HaeIII,13-HinfI);14–16:Lactobacillusharbinensis(14-MspI,15-HaeIII,16-HinfI);17–19:Lactobacillusparabuchneri profile(17-MspI,18-HaeIII,19-HinfI).
10000 bp
500 bp 1000 bp
250 bp
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Fig.2–Digestionprofileoflacticacidbacteriaisolatedfromcassavafermentationinacassavaflourmanufacturer.Lanes:1: 1kbDNAladder;2–4:Lactobacillusghanensisprofile(2-MspI,3-HaeIII,4-HinfI);5–7:Enterococcusfaeciumprofile(5-MspI, 6-HaeIII,7-HinfI);8–10:Weisellacibariaprofile(8-MspI,9-HaeIII,10-HinfI);11–13:Lactococcusgarvieaeprofile(11-MspI, 12-HaeIII,13-HinfI);14–16:Lactobacilluslactisprofile(14-MspI,15-HaeIII,16-HinfI);17–19:Leuconostocmesenteroidesprofile (17-MspI,18-HaeIII,19-HinfI).
1 10000 bp 1500 bp 1000 bp 750 bp 2 3 4 5 6 7 8 9 10 11 12
Fig.3–PCRfingerprintingprofileofyeastsisolatedfrom cassavafermentationinacassavaflourmanufacturerin Formiga(MG,Brazil).Lanes:1:1kbDNAladder;2,5–7:
Kazachstaniaexigua;3,4and8:Candidahumilis;9and10:
Pichiascutulata;11:Candidaethanolica;12:Geotrichum
fragrans.
L.plantarumwheninassociationwiththeyeast.Onthe21st day,thepresenceoftheyeastwasnolongerobserved(Fig.5). Cassavastarch,thedriedfinalproduct,obtainedwithL. plantarumexhibitedthehighestTTAvalueamongthesamples.
Table1–TTAandpHofsamplesfromacassavaflour
manufacturerinthemunicipalityofFormiga,Minas
Gerais(MG)state,Brazil.
Time TTA pH
19 3.31 ±0.17 c 4.01 ±0.01 a
26 6.83 ±0.25 b 3.75 ±0.02 b
33 9.36 ±0.20 a 3.53 ±0.07 cd
40 9.77 ±0.23 a 3.46 ±0.02 d
Meanvalues±standarddeviationinthesamecolumnfollowedby differentsuperscriptlettersaresignificantlydifferent(p<0.001).
Althoughthisvaluewasnotclosetothevalueofcommercial product, it waswithinthe establishedlimits oflegislation,
whichmandates amaximumacidityof1.0%forsweet
cas-savastarchandamaximumacidityof5.0%forsourcassava starch.32Hence,theproductobtainedwithL.plantarum
sin-gleculturecouldbecalledsourcassavastarch.Finalproducts obtainedwith other starter culturesmightbe calledsweet cassavastarch.ThepHvalueofstarchobtainedwithL. plan-tarum singleculturewas the onlyvaluethat didnotdiffer fromthepHvalueofthecommercialproduct.Theexpansion capacitiesforthecassavastarchesobtainedwiththestarter
Table2–TTAafter24and48hoffermentationwith
lacticacidbacteriaisolatedfromacassavaflour
manufacturerinthemunicipalityofFormiga,Minas
Gerais(MG)state,Brazil.
Species TTA(%)
24h 48h
Lb.brevis 3.47 ±0.06 ay 5.28 ±0.07 ax
Lc.garvieae 3.37 ±0.20 ay 5.28 ±0.14 ax
Lb.plantarum 2.96 ±0.08 ay 4.97 ±0.19 abx
Lc.lactissubsp.lactis 1.63 ±0.09 ay 4.47 ±0.24 abx
E.faecium 2.79 ±0.15 ay 4.35 ±0.10 abx Ln.mesenteroides 2.74 ±0.44 ay 3.77 ±0.03 abx Lb.parabuchneri 2.07 ±0.08 ay 3.51 ±0.10 abx Lb.fermentum 2.57 ±0.09 ay 3.50 ±0.03 abx W.cibaria 0.97 ±0.16 ay 3.11 ±0.05 abx Lb.ghanensis 0.62 ±0.13 ay 2.68 ±0.14 abx Lb.harbinensis 2.59 ±0.04 ax 2.55 ±0.10 abx Lb.casei/Lb.paracasei 1.26 ±0.04 ay 2.33 ±0.20 bx
Meanvalues±standarddeviationinthesamelinefollowedby dif-ferentsuperscriptletters(x,y)aresignificantlydifferent(p<0.001). Meanvalues±standarddeviationinthesamecolumnfollowedby differentsuperscriptletter(a,b)aresignificantlydifferent(p<0.001).
Table3–TTAafter24and48hoffermentationwith
yeastsisolatedfromacassavaflourmanufacturerinthe
municipalityofFormiga,MinasGerais(MG)state,Brazil.
Species TTA(%) 24h 48h Geotrichumfragrans 1.27 ±0.00 ay 1.60 ±0.10 ax Pichiascutulata 0.48 ±0.10 ay 0.78 ±0.05 ax Candidaethanolica 0.39 ±0.05 ax 0.45 ±0.00 ax Candidahumilis 1.03 ±0.14 ax 0.81 ±0.00 ay Kazachstaniaexigua 1.18 ±0.09 ax 0.94 ±0.05 ay
Meanvalues±standarddeviationinthesamelinefollowedby dif-ferentsuperscriptletters(x,y)aresignificantlydifferent(p<0.001). Meanvalues±standarddeviationinthe samecolumn followed bydifferent superscriptletter(a) arenot significantly different (p>0.05).
culturesandforcommercialproductdidnotdiffer(Table4). Cassavastarchesproducedviaapilot-scalefermentation pro-cessprovedtobewithinthemicrobiologicallimits31andwere
thereforesuitableforhumanconsumption.
Discussion
DuringtheisolationofLABandyeastsfromcassava fermen-tation,LABwerepresentinallsamplescollectedwithhigher counts than those obtained foryeasts. Many authors also reportedthepresenceofLABasprevalentmicroorganismsin cassavafermentations.33–35HighLABcountsareevidenceof
theimportanceofthesemicroorganismsincassava fermen-tation.
AmongtheLABoftenfoundintraditionalcassava fermen-tations,manyauthorsreportthepresenceofL.brevisandL. plantarum,7,9,10,34,36,37 whichwere amongthemostfrequent
isolatesinthepresentstudy.Lacerdaetal.6reportedL.
plan-tarumandL.fermentumasthepredominantLABduringsour
cassavastarchfermentationisolatedfromtwocassavaflour manufacturerslocatedinthemunicipalityofConceic¸ãodos Ouros(MG,Brazil).L.breviswasisolatedinminorproportions. Thecountsofthesebacteriaweresimilartothosefoundin thisstudy.Themicrobiotafromcassavafermentationhas dif-ferentoriginsandmaycomefromrawmaterials,utensilsand equipmentusedinitsproduction.Insectsorhandlerscanalso carrythesemicroorganisms.
Themostcommonlyisolatedyeastspecies,P.scutulata,K. exigua,C.humilis,G.fragransandC.ethanolicawerealsofound inothercassavafermentationstudies.Lacerdaetal.6collected
samplesfromsourcassavastarchfermentationandidentified thefollowingyeastspecies:Pichiasp.,genusthatwasformerly knownasIssatchenkiasp.,wasfoundincountsof5log10CFU/g;
C.humilisandC.ethanolicawerealsofrequentlyisolated. Wil-fridPadanou10foundaprevalenceofSaccharomycescerevisiae
(22%)and P.scutulata(20%)inthe fermentationoflafun,an Africanproductobtainedfromcassavasubmerged fermenta-tion.Theoccurrenceoftheseyeastssuggeststhattheycould contributetothe sensorialquality. Severalstudies revealed thepresenceofthegenusCandidaincassavafermentationsto obtaindifferentproducts.6,7,10,38,39Inthepresentstudy,one
speciesofthegenusPichiaandtwospeciesofthegenus Can-didawereisolated,showingthatspeciesofthesegeneracan growinacidicconditionsfoundincassavafermentationtanks. TheacidicenvironmentcreatedbyLABfavorsthe prolif-erationofyeastsinfoods.Simultaneously,thegrowthofLAB isstimulatedbythepresenceofyeasts,whichprovidegrowth factorssuchasvitaminsandnitrogencompounds.40The
asso-ciationofLABandyeastsduringfermentationhasasignificant impactonfoodqualityparameterssuchastexture,flavorand nutritional values.41 Even though the complex interactions
between LABand yeastsare not yetfully understood,it is knownthatLABandyeastshavetheabilitytoadapttomany differentsubstrates.42
Thedecrease ofpHvaluesand theconsequent increase in TTA of samples from cassava flour manufacturer were expectedresultsbecause,duringfermentation,thesynthesis oforganicacidsoccurs,leadingtoacidificationofthemedium. Itispossiblethatyeastshaveasmallcontributionordonot contributetoacidification,especiallybecausemostyeastsare not resistant to extreme acidic conditions. The pH values foundinthepresentstudyarelowerthanthosedetermined byCoulinetal.7inattiéké,afermentedcassavaproduct.The
averagepHofthisfoodduringthetraditionalfermentationin smallscalewas5.0.
According to the microbiological analyses, sour cassava starch from the fermentation tank studied would be con-sideredsuitableforhumanconsumption.Verificationofthe presenceofB.cereusisimportantbecauseitproducestoxins andmaycausefoodpoisoningwhenconsumedinnumbers higherthan105CFU/g.Besidestheabilitytoproduceorganic
acids,LABalsocontributetoinhibitionofpathogensdueto productionofantimicrobialcompoundssuchashydrogen per-oxide,diacetylandbacteriocins.43,44LABacidificationpoweris knowntobeeffectiveincontrollinggrowthofmicroorganisms infoodandextendingitsshelflife.45
The microorganisms considered as predominant in the presentstudyweretheoneswithhighercountsandthemost frequentlyisolatedduringcassavafermentationprocess.The
1,20 6,00 5,00 4,00 3,00 2,00 1,00 0,00 1,00 0,80 0,60 0,40 0,20 0,00 0 7
Time (days) Time (days)
TT
A (%) pH
14 21 28 0 7 14 21 28
Fig.4–MonitoringofTTAandpHduringcassavapilot-scalefermentationwithselectedstartercultures.
8 7 6 5 4 3 2 1 0 0 7 14
Time (days) Time (days)
Time (days) Time (days)
Count (log 10 CFU/g) 8 7 6 5 4 3 2 1 0 Count (log 10 CFU/g) 8 7 6 5 4 3 2 1 0 Count (log 10 CFU/g) 8 7 6 5 4 3 2 1 0 Count (log 10 CFU/g) 21 28 0 7 14 21 28 0 7 14 21 28 0 7 14 21 28
Fig.5–Counts(log10CFU/g)oftheselectedstarterculturesduringcassavapilot-scalefermentation.
Table4–TTA,pHandexpansioncapacityforcassavastarches(driedfinalproducts)obtainedinapilotscale
fermentationcomparedwithacommercialproduct.
Cassavastarches TTA(%) pH Expansioncapacity
Lactobacillusbrevis 0.57 ±0.06 cd 6.07 ±0.04 a 1.04 ±0.00 a
Lactobacillusplantarum 1.41 ±0.00 b 3.64 ±0.03 d 0.98 ±0.04 a
L.brevis+Pichiascutulata 0.71 ±0.10 c 4.34 ±0.06 c 1.10 ±0.08 a
L.plantarum+P.scutulata 0.40 ±0.00 d 4.84 ±0.02 b 1.05 ±0.02 a
Commercialproduct 6.75 ±0.20 a 3.62 ±0.04 d 1.06 ±0.05 a
Meanvalues±standarddeviationinthesamecolumnfollowedbydifferentsuperscriptlettersaresignificantlydifferent(p<0.001).
resultsare inagreementwith the findingsofpredominant
microorganismsintraditionalcassavafermentations,
includ-ing sourcassava starchfermentation, asreported inmany
previousstudies.6–11,33–38,46–48Therefore,themicroorganisms
isolated and identified inthe present study are a suitable
sourceofmicroorganismsforuseintestsfortheselectionof startercultures.
Selection of starter cultures showed that some yeast species exhibitedamylolytic activity. Thisobservation sug-gests that yeasts possibly play an important role at the
beginning of fermentation, degrading starch and releasing sugarsforthe growthofLABandtheir own.Ialsoshowed that TTA values obtained by LAB were higher than those obtainedbyyeasts,whichindicatesthattheformeraremainly responsible for acidification, which was also confirmed by OguntoyinboandDodd.35
L.plantarum,selectedasastarterculturefortheproduction ofsourcassavastarchinpilot-scalefermentationprocess,has alreadybeen evaluatedasastartercultureforthe fermen-tationofothertraditionalcassavaproduct.Kostineketal.33
assessedthe technological properties ofthe prevalent LAB inthefermentationofgari,anAfrican productobtainedby cassavasolid-statefermentation,andnotedthatL.plantarum
exhibitedbetterand fasteracid productionamongthe iso-latedbacteriaandwasrecommendedasthestarterculture. Huchetal.49testedL.plantarumandconcludedthatitssuccess
to predominatein cassava fermentation demonstrates the potentialforitsdevelopmentasastartercultureforgari indus-trialization.Edwardetal.50investigatedtheuseoflyophilized
LABstrainsasstarterculturesforgariproductionandfound thatL.plantarumcouldbeproducedatlowcost.
MonitoringofTTArevealed valueshigherthanthe ones determinedbyVogelmannetal.42whenstudyingcassava
fer-mentedbyanassociationofLABandyeastsusedasstarter cultures.Thevaluesfoundbytheauthorsrangedfrom10.3to 11.5SH(DegreesSoxhlet-Henkel),whichcorrespondsto0.23% and0.26%,respectively.MonitoringofpHrevealedvalues simi-lartotheonesdeterminedbyVogelmannetal.,42whichvaried from4.1to4.2duringthe12daysoffermentation.
Themonitoringoftheviabilityofstarterculturesshowed that the culturesused were able toprevail duringcassava pilot-scale fermentation. This result may suggest a strong associationbetweenLABandyeastsatthebeginningof fer-mentation and indicates that LAB play an important role duringthe entireprocess,especiallyconcerningthequality andsafetyofthefinalproduct.Besidesthat,theuseofstarter culturesmay optimizethefermentationprocessdecreasing thefermentationtime.Huchetal.49testedL.plantarumand
L.fermentumasstarterculturesfortheproductionofgari.The molecularmonitoringbyRandomAmplifiedPolymorphicDNA (RAPD)-PCRand typingtechniquesbyPulsedFieldGel Elec-trophoresis(PFGE)indicatedthatL.plantarumwassuccessful inassertingitselfasapredominantstrain,whichdidnot hap-penwithL.fermentum.
L.plantarumshowedbetterperformanceasastarterculture forsourcassavastarchproductioninapilot-scale fermenta-tionprocess.Cassavastarchproducedbythisstrainwasfound tohavethehighestvalueofTTAwhencomparedwiththe otherstartercultures;the pHandexpansioncapacitywere not different from the valuesobtained by the commercial product;also,therewasnoevidenceofpathogens.Therefore, furtherstudiesarerequiredfortheestablishmentofastarter culturethatwillcontributetothestandardizationofthe cas-savafermentationconditions,therebyensuringhigherquality productsandconsumeracceptability.
Conflicts
of
interest
Theauthorsdeclarenoconflictsofinterest.
r
e
f
e
r
e
n
c
e
s
1.HowelerR,LutaladioN,ThomasG.TitleofSubordinate Document.SaveandGrow:Cassava:AGuidetoSustainable ProductionIntensification.Rome:FoodandAgriculture OrganizationoftheUnitedNations;2013.
http://www.fao.org/docrep/018/i3278e/i3278e.pdf.
2.IBGE,ftp://ftp.ibge.gov.br/ProducaoAgricola/
Levantamento SistematicodaProducaoAgricola
[mensal/Comentarios/lspa201403comentarios.pdf]Titleof
subordinatedocument.Indicadores:produc¸ãoagrícola: levantamentosistemáticodaproduc¸ãoagrícola:marc¸ode2014;
2014.
3.CookeRD,CourseyDG.Cassava:AMajorCyanide-Containing
FoodCrop.CyanideinBiology.NewYork:AcademicPress; 1981:93–114.
4.NoutMJR,SarkarPK.Lacticacidfoodfermentationin
tropicalclimates.AntonievanLeeuwenhoek.
1999;76(1–4):395–401.
5.FranzCMAP,HuchM,MatharaJM,etal.Africanfermented
foodsandprobiotics.IntJFoodMicrobiol.2014;190:84–96.
6.LacerdaICA,MirandaRL,BorelliBM,etal.Lacticacid
bacteriaandyeastsassociatedwithspontaneous
fermentationsduringtheproductionofsourcassavastarch
inBrazil.IntJFoodMicrobiol.2005;105(2):213–219.
7.CoulinP,FarahZ,AssanvoJ,SpillmannH,PuhanZ.
Characterisationofthemicrofloraofattiéké,afermented
cassavaproduct,duringtraditionalsmall-scalepreparation.
IntJFoodMicrobiol.2006;106(2):131–136.
8.CrispimSM,NascimentoAMA,CostaPS,etal.Molecular
identificationofLactobacillusspp.associatedwithpuba,a
Brazilianfermentedcassavafood.BrazJMicrobiol.
2013;44(1):15–21.
9.KostinekM,SpechtI,EdwardVA,etal.Characterisationand
biochemicalpropertiesofpredominantlacticacidbacteria
fromfermentingcassavaforselectionasstartercultures.Int
JFoodMicrobiol.2007;114(3):342–351.
10.WilfridPadonouS,NielsenDS,HounhouiganJD,ThorsenL,
NagoMC,JakobsenM.ThemicrobiotaofLafun,anAfrican
traditionalcassavafoodproduct.IntJFoodMicrobiol.
2009;133(1–2):22–30.
11.LacerdaICA,GomesFCO,BorelliBM,etal.Identificationof
thebacterialcommunityresponsiblefortraditional
fermentationduringsourcassavastarch,cachac¸aandminas
cheeseproductionusingculture-independent16srRNAgene
sequenceanalysis.BrazJMicrobiol.2011;42:650–657.
12.AidooKE,NoutMJ,SarkarPK.Occurrenceandfunctionof
yeastsinAsianindigenousfermentedfoods.FEMSYeastRes.
2006;6(1):30–39.
13.CapliceE,FitzgeraldGF.Foodfermentations:roleof
microorganismsinfoodproductionandpreservation.IntJ
FoodMicrobiol.1999;50(1):131–149.
14.AtrihA,RekhifN,MoirAJG,LebrihiA,LefebvreG.Modeof
action,purificationandaminoacidsequenceofplantaricin
C19,ananti-ListeriabacteriocinproducedbyLactobacillus
plantarumC19.IntJFoodMicrobiol.2001;68(1):93–104.
15.MessensW,DeVuystL.Inhibitorysubstancesproducedby
Lactobacilliisolatedfromsourdoughs—areview.IntJFood
Microbiol.2002;72(1–2):31–43.
16.MiduraTF,BryantRG.Samplingplans,samplecollection,
shipment,andprepartionsforanalysis.In:DownesFP,ItoK,
eds.CompendiumofMethodsfortheMicrobiologicalExamination ofFoods.4thed.Washington:AmericanPublicHealth
Association(APHA);2001:13–23.Chapter2.
17.NorrisJR,BerkeleyRCW,LoganNA,O’donnellAG.Thegenera
BacillusandSporolactobacillus.In:StarrMP,StolpH,Truper
onHabitats,IsolationandIdentificationofBacteria.
1981:1711–1742.
18.HoffmanCS,WinstonF.Aten-minuteDNApreparationfrom
yeastefficientlyreleasesautonomousplasmidsfor
transformaionofEscherichiacoli.Gene.1987;57(2–3):267–272.
19.LaneDJ.16S/23SrRNAsequencing.In:StackebrandtE,
GoodfellowM,eds.NucleicAcidTechniquesinBacterial
Systematics.NewYork:Wiley;1991:115–175.
20.BrightwellG,BoeremaJ,MillsJ,MowatE,PulfordD.
Identifyingthebacterialcommunityonthesurfaceof
IntraloxTMbeltinginameatboningroomby
culture-dependentandculture-independent16SrDNA
sequenceanalysis.IntJFoodMicrobiol.2006;109(1):47–53.
21.TamuraK,StecherG,PetersonD,FilipskiA,KumarS.MEGA6:
molecularevolutionarygeneticsanalysisversion6.0.MolBiol
Evol.2013;30(12):2725–2729.
22.TorrianiS,FelisGE,DellaglioF.DifferentiationofLactobacillus
plantarum,L.pentosus,andL.paraplantarumbyrecAgene
sequenceanalysisandmultiplexPCRassaywithrecA
gene-derivedprimers.ApplEnvironMicrobiol.
2001;67:3450–3454.
23.daSilva-FilhoEA,DosSantosSKB,doMonteResendeA,de
MoraisJOF,deMoraisMAJr,SimõesDA.Yeastpopulation
dynamicsofindustrialfuel-ethanolfermentationprocess
assessedbyPCR-fingerprinting.AntonievanLeeuwenhoek.
2005;88(1):13–23.
24.LieckfeldtE,MeyerW,BörnerT.Rapididentificationand
differentiationofyeastsbyDNAandPCRfingerprinting.J
BasicMicrob.1993;33(6):413–425.
25.LachanceM-A,BowlesJM,StarmerWT,BarkerJSF.
KodamaeakakaduensisandCandidatolerans,twonew
ascomycetousyeastspeciesfromAustralianHibiscus
flowers.CanJMicrobiol.1999;45(2):172–177.
26.OfficialMethodsofAnalysisofAOACINTERNATIONAL.20thEd. Gaithersburg,MD,USA:AOACINTERNATIONAL;2016.
www.eoma.aoac.org.
27.BennettRW,BelayN.Bacilluscereus.In:DownesFP,ItoK,
eds.CompendiumofMethodsfortheMicrobiologicalExamination ofFoods.Washington,DC:AmericanPublicHealth
Association;2001:311–316.
28.SanniAI,Morlon-GuyotJ,GuyotJP.Newefficient
amylase-producingstrainsofLactobacillusplantarumandL.
fermentumisolatedfromdifferentNigeriantraditional
fermentedfoods.IntJFoodMicrobiol.2002;72(1–2):53–62.
29.YarrowD.Methodsfortheisolation,maintenance
classificationandidentificationofyeasts.In:KurtzmanCP,
FellJW,eds.TheYeasts:ATaxonomicStudy.Amsterdam:
Elsevier;1998:77–100.
30.MaedaKC,CeredaMP.Avaliac¸ãodeduasmetodologiasde
expansãoaofornodopolvilhoazedo.FoodSciTechnol.
2001;21(2):139–143.
31.Brazil.MinistériodaSaúde.Resoluc¸ãoRDCn◦12,de02de
janeirode2001.Aprovaoregulamentotécnicosobreospadrões microbiológicosparaalimentos.DiárioOficialdaUnião,sec¸ão1;
2001.
32.Brazil.Decreton◦12.486,de20deoutubrode1978.Aprova
normastécnicasespeciaisrelativasaalimentosebebidas.Diário
OficialdoEstadodeSãoPaulo;1978:3–25.
33.KostinekM,SpechtI,EdwardVA,etal.Diversityand
technologicalpropertiesofpredominantlacticacidbacteria
fromfermentedcassavausedforthepreparationofGari,a
traditionalAfricanfood.SystApplMicrobiol.
2005;28(6):527–540.
34.ObilieEM,Tano-DebrahK,Amoa-AwuaWK.Souringand
breakdownofcyanogenicglucosidesduringtheprocessing
ofcassavaintoakyeke.IntJFoodMicrobiol.2004;93(1):115–121.
35.OguntoyinboFA,DoddCER.Bacterialdynamicsduringthe
spontaneousfermentationofcassavadoughingari
production.FoodControl.2010;21(3):306–312.
36.KimaryoVM,MassaweGA,OlasupoNA,HolzapfelWH.The
useofastartercultureinthefermentationofcassavaforthe
productionof“kivunde”,atraditionalTanzanianfood
product.IntJFoodMicrobiol.2000;56(2–3):179–190.
37.OmarNB,AmpeF,RaimbaultM,GuyotJ-P,TailliezP.
MoleculardiversityoflacticacidbacteriafromCassavasour
starch(Colombia).SystApplMicrobiol.2000;23(2):285–291.
38.OguntoyinboFA.EvaluationofdiversityofCandidaspecies
isolatedfromfermentedcassavaduringtraditionalsmall
scalegariproductioninNigeria.FoodControl.
2008;19(5):465–469.
39.OyewoleOB.Characteristicsandsignificanceofyeasts’
involvementincassavafermentationfor‘fufu’production.
IntJFoodMicrobiol.2001;65(3):213–218.
40.NoutMJR.Ecologyofacceleratednaturallacticfermentation
ofsorghum-basedinfantfoodformulas.IntJFoodMicrobiol.
1991;12(2):217–224.
41.BoonnopK,WanapatM,NontasoN,WanapatS.Enriching
nutritivevalueofcassavarootbyyeastfermentation.Sci
Agric.2009;66(5):629–633.
42.VogelmannSA,SeitterM,SingerU,BrandtMJ,HertelC.
Adaptabilityoflacticacidbacteriaandyeaststosourdoughs
preparedfromcereals,pseudocerealsandcassavaanduseof
competitivestrainsasstarters.IntJFoodMicrobiol.
2009;130(3):205–212.
43.GálvezA,AbriouelH,LópezRL,OmarNB.Bacteriocin-based
strategiesforfoodbiopreservation.IntJFoodMicrobiol.
2007;120(1–2):51–70.
44.NardiRMD,SantoroMM,OliveiraJS,etal.Purificationand
molecularcharacterizationofantibacterialcompounds
producedbyLactobacillusmurinusstrainL1.JApplMicrobiol.
2005;99(3):649–656.
45.BealesN.Adaptationofmicroorganismstocold
temperatures,weakacidpreservatives,lowpH,andosmotic
stress:areview.ComprRevFoodSciFoodSaf.2004;3(1):1–20.
46.ManteES,Sakyi-DawsonE,Amoa-AwuaWK.Antimicrobial
interactionsofmicrobialspeciesinvolvedinthe
fermentationofcassavadoughintoagbelimawithparticular
referencetotheinhibitoryeffectoflacticacidbacteriaon
entericpathogens.IntJFoodMicrobiol.2003;89(1):41–50.
47.MugulaJK,NarvhusJA,SørhaugT.Useofstarterculturesof
lacticacidbacteriaandyeastsinthepreparationoftogwa,a
Tanzanianfermentedfood.IntJFoodMicrobiol.
2003;83(3):307–318.
48.ObadinaAO,OyewoleOB,SanniLO,TomlinsKI.
Bio-preservativeactivitiesofLactobacillusplantarumstrainsin
fermentingCasssava‘fufu’.AfrJBiotechnol.2006;5:620–623.
49.HuchM,HanakA,SpechtI,etal.UseofLactobacillusstrains
tostartcassavafermentationsforGariproduction.IntJFood
Microbiol.2008;128(2):258–267.
50.EdwardVA,HuchM,DortuC,etal.Biomassproductionand
small-scaletestingoffreeze-driedlacticacidbacteriastarter
strainsforcassavafermentations.FoodControl.